为了大幅度提高水印方法的鲁棒性,提出了一种视频语义水印方法,利用高级语义稳定且不易受攻击的特点,将视频语义与水印相结合.该方法将关联规则挖掘看作约束优化问题,使用改进的克隆选择算法进行模糊关联规则挖掘,以提取运动语义和纹理语义,并在线生成动态视频语义水印;根据运动语义自适应确定感兴趣镜头,根据纹理语义自适应确定感兴趣Ⅰ帧,根据人眼视觉掩蔽特性,选择运动剧烈和运动缓慢区域作为感兴趣区域,将水印嵌入在感兴趣Ⅰ帧的亮度子块预测残差离散余弦变换中频系数上;利用视频纹理特征,自适应控制水印嵌入强度.实验和分析表明,该方法不仅对各种常规攻击鲁棒,而且对帧重组、帧内裁剪和帧删除等视频特有攻击表现出强的鲁棒性.%To improve the robustness of the watermarking method remarkably, this paper proposes a video semantics watermarking method, which combines video semantics with video watermarking by using the high-level semantic features of stability and unvulnerability. This method looks upon mining association rules as a constrained optimization problem, uses an improved clonal selection algorithm for the fuzzy association rule mining to extract the semantics of motion and texture, and generates dynamic semantics watermark online. This method determines the shots of interest according to motion semantics adaptively, determines the I frames of interest according to texture semantics adaptively, and selects the intense movement and slow movement of the region as a region of interest according to human visual masking properties, and then embeds watermark in the IF DCT coefficients of the luminance sub-block prediction residual of the I frames of interest; it adaptively controls watermark embedding strength by use of video texture features. Experiments and analysis show that this method is robust not only to various conventional attacks, but also to re-frame, frame cropping, frame deletion and other video-specific attacks.
展开▼
